The growth rates of Antarctic fur seal Arctocephalus gazella pups estimated from weighing cross-sections of the population were compared with measured/inferred changes in the the availability of their main prey species, Antarctic krill Euphausia superba from 1989-2000. There was no relationship between growth rate and mass at weaning and there were counterintuitive indications of higher growth rates in years of low krill availability. Biases reflecting changes in the component of the population available for sampling appear to invalidate the widely held assumption that inter-annual differences in growth rate can reliably be derived from differences in the slope of a linear relationship based on cross-sectional population samples. A new index was developed, based on the deviation of pup mass at age in each year compared to the multi-year mean, that was not dependent on assumptions of linearity. The indices of growth deviates produced a more logical relationship with other indices of pup development and related more appropriately to variations in prey availability. The potential impact of methodological biases on the interpretation of growth rate suggests that comparisons of growth rates should not rely on assumptions regarding the underlying growth pattern.